Study objectives: To determine the predictive value of abnormalities on high-resolution CT (HRCT) on pulmonary disease in systemic sclerosis.
Patients: Fifty-two patients suffering from systemic sclerosis.
Design: Pulmonary disease was defined by pulmonary function test abnormalities, ie, total lung capacity (TLC) <80% of predicted value and/or diffusion of carbon monoxide (DLCO) <75% of predicted value, without any pulmonary event other than systemic sclerosis in the medical history. Patients were divided in two groups, group A with pulmonary disease (29 patients) and group B without pulmonary disease (23 patients). HRCT abnormalities were scored on whole lungs. A decision matrix was constructed to determine sensitivity, specificity, positive and negative predictive values, and false-positive and false-negative rates. A receiver operating characteristic curve was constructed to determine the best compromise between sensitivity and specificity.
Results: HRCT total scores were higher in group A (9.0+/-4.3) than in group B (5.0+/-2.8) (p < 0.001) and they correlated with TLC (r =-0.39, p < 0.005) and DLCO (r = -0.50, p < 0.0002). An HRCT score of 7 corresponded to the best compromise between sensitivity (0.60) and specificity (0.83), with a positive predictive value of 0.82. Taking into account a value of 10 for the HRCT score increased specificity to 1 but decreased sensitivity to 0.41.
Conclusion: A minimum score of 7 would be required to consider HRCT abnormalities in systemic sclerosis as predictive of pulmonary disease. An HRCT score of 10 makes it possible to establish the diagnosis of lung involvement severe enough to impair pulmonary function.